BackendForge: leaderboard
BackendForge (Peking University, 2026) asks coding agents to build a complete Dockerized Python backend service from a written specification and an OpenAPI contract for 56 services rewritten from open-source applications, judged only through black-box HTTP tests. Task success rate in percent, where a task counts only when the service passes every item of the final test oracle, including reviewed tests added during oracle co-evolution; higher is better.
Metric: Task success rate (%; final oracle of 7,890 items including 640 reviewed co-evolved items; 56 contract-defined backend services rewritten from open-source applications; from a visible spec and OpenAPI contract the agent builds a Dockerized Python service, scored only through black-box HTTP pytest items; a task counts only when every item passes; one attempt per task in the same mini-SWE-style coding harness). Source: arxiv.org. Saturation forecast: Around February 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 (xHigh) | 28.57 | #49 (GPT-5.5) |
| 2 | Claude Opus 4.7 (Max) | 17.86 | #79 (Claude Opus 4.7) |
| 3 | Gemini 3.5 Flash | 5.36 | #101 |
| 4 | GLM-5.1 | 3.57 | #179 |
| 5 | DeepSeek V4 Pro (Max) | 3.57 | #188 (DeepSeek V4 Pro) |
| 6 | Claude Sonnet 4.6 (Max) | 3.57 | #161 (Claude Sonnet 4.6) |
| 7 | Kimi K2.6 | 1.79 | #178 |
| 8 | DeepSeek V4 Pro (High) | 1.79 | #188 (DeepSeek V4 Pro) |
| 9 | MiniMax-M2.7 | 0 | #518 |
| 10 | DeepSeek V4 Flash (Max) | 0 | #258 (DeepSeek V4 Flash) |
| 11 | DeepSeek V4 Flash (High) | 0 | #258 (DeepSeek V4 Flash) |
| 12 | Hy3-preview (High) | 0 | #316 (Hy3-preview) |
| 13 | Qwen 3.6 Plus (High) | 0 | #222 (Qwen 3.6 Plus) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=backendforge · How It Works · Data refreshed daily, snapshot 2026-11-13.